It’s time to nail down the Data Center as a Service (DCaaS) definition
It’s no secret that the COVID-19 pandemic has accelerated digital transformation around the world. According to research firm Information Services Group, the pandemic has sped up digital transformation by three to five years as enterprises seek to increase operational efficiency, improve customer experiences, and boost competitiveness. For enterprises that have been cautious about adopting digital transformation and possess limited expertise in cloud computing, Data Center as a Service (DCaaS) can offer a proven, ready-to-deploy solution for cloud enablement. However, it can be confusing to try to figure out exactly what DCaaS is and what benefits it provides, as solutions with fundamentally different operating models are branded as DCaaS. This begs the question: What is today's Data Center as a Service definition for enterprises?
The evolving Data Center as a Service definition
If you search the Internet for “Data Center as a Service,” abbreviated DCaaS, you will likely find information from managed hosting providers or colocation data centers. These companies rent or lease facility space and sometimes equipment to businesses that don’t want to rack and stack their own physical hardware, often because they lack sufficient IT staff, budget or on-prem data center space. These business models date back to the 1990s, during the dot com boom; the original meaning of “Data Center as a Service” also emerged from that era.
From the mid-1990s to the early 2010s, “Data Center as a Service'' referred to a service that not only rented or leased facility space for clients’ hardware, but also included hardware deployment services. Basically, it was co-location with the bonus of installation assistance.
However, a few things happened in the computing world. In the 2000s, with the popularization of cloud computing and a burgeoning market for web-hosted applications, the software-as-a-service (SaaS) paradigm really took off. Around the same time, as demand for SaaS fueled the need for greater computing power, the virtualization market also began to flourish, spreading the concept of software abstraction of hardware resources and functionality. All layers of the software stack became targets for software abstraction, and much of the hardware ecosystem as well. Soon we encountered a plethora of “as-a-service” categories and acronyms, ranging from infrastructure-as-a-service (IaaS), platform-as-a-service (PaaS), and bare metal-as-a-service (BMaaS), to firewall-as-a-service (FWaaS), load balancing-as-a-service (LBaaS), and even more obscure examples like Field Programmable Gate Array-as-a-service (FPGAaaS).
What all these “as-a-service” offerings have in common is their ability to provide software-defined functionality on commodity hardware, replacing costly proprietary physical appliances. Many solutions are based on open source projects, such as OpenStack for IaaS and NGINX for LBaaS.
Thus, in the dot com era, “Data Center as a Service” offerings were primarily about outsourcing the physical real estate for housing company-owned servers, and networking and storage appliances. Colos provided a protected facility space, power, cooling, and a secure network connection, often near major internet exchanges. Nowadays, “as a service” technologies also enable businesses to offload physical IT needs, but these needs are fundamentally delivered through software abstraction of hardware capabilities. Today, the Data Center as a Service definition revolves around a software-defined data center, which provides core data center functionality through software abstraction of compute, networking, and storage resources.
Today’s Data Center as a Service requirements
A number of enterprise IT trends influence the shape of DCaaS offerings today. According to the Flexera 2021 State of the Cloud Report, 82% of respondents are taking a hybrid cloud approach in their enterprise cloud strategy. Additionally, containers have become mainstream, with 53% of respondents already using Docker and 48% utilizing Kubernetes.
Although the Flexera report showed that three-fourths of enterprises use or plan to use containers, today most workloads still run in VMs. According to a Virtualization and Containerization Survey conducted by Mirantis in 2021, 71% of 667 respondents ran the majority of their workloads in virtual machines rather than containers.
Given these various trends, Data Center as a Service solutions today must enable enterprises to leverage the power of software abstraction to seamlessly deploy and manage resources for both virtualized and containerized workloads across multiple public and private clouds. According to the Flexera report, 92% of enterprise respondents have a multi-cloud strategy, with 82% taking a hybrid approach. As more enterprises mandate digital transformation, modern DCaaS providers need to enable both workload portability across diverse cloud infrastructures and a painless transition from legacy virtualized to modern containerized workloads.
Like DCaaS offerings of the dot com era, software-defined DCaaS solutions today enable businesses to outsource physical aspects of IT infrastructure. However, instead of choosing from only a limited set of locations from a specific DCaaS provider, companies can now choose to easily leverage heterogeneous resources located on-premises, in colo/managed hosting facilities, as well as in public clouds — all unified by a single hybrid cloud platform.
This seamless experience can be made possible through a single API managing access across cloud infrastructure providers no matter where clusters are deployed. A single pane of glass across public and private clouds and across virtualized and containerized resources also enables holistic monitoring of disparate systems and resources.
Importantly, as enterprises increasingly adopt open source strategies to eliminate vendor lock-in, reduce costs, and stay on the bleeding edge of innovation, Data Center as a Service platforms should leverage best-of-breed open source components. According to BCG analysis of GitHub data, 99% of Fortune 500 companies currently use open source software.
DCaaS is not just about software
Although software-defined functionality is core to DCaaS solutions today, it’s not all about the software. The “as a service” part of “Data Center as a Service” also refers to comprehensive services and support included with DCaaS offerings. More than ever, enterprises today face enormous pressure to achieve faster time to production and shorter release cycles.
At the same time, one of the top challenges that enterprises face is the lack of in-house expertise to effectively leverage complex cloud and container technologies. For companies everywhere, hiring and retaining cloud and container expertise is extremely challenging and presents a risk to business continuity. The pandemic-induced Great Resignation in the last two years has only worsened the situation.
For these reasons, many enterprises today need fully managed Data Center as a Service so they can hand off the burden of staffing and operating a cloud native data center, and focus precious resources on driving greater innovation and contributing business value. With a fully managed Data Center as a Service, developers can finally devote their time and energy to developing apps, and enterprises can concentrate on their core competency. It’s known as a “zero-ops” approach to IT.
For companies that cannot hand over full control of data center operations, a “co-pilot” approach can be adopted. This approach should provide proactive 24x7x365 support by cloud and container experts, who can provide indispensable guidance on troubleshooting issues and maintaining the latest cloud native technologies.
White glove service from the start
Digital transformation can be daunting without the proper expertise and guidance. Data center as a service providers should give businesses all the necessary services to quickly get customers up and running. Customers typically need help with hardware configuration, software installation, cluster deployment and validation, workload selection and onboarding, and the DCaaS solution should provide that assistance.
Additionally, in the course of normal data center operations, a lot of administrative work is required to manage third-party hardware vendors and service providers. A comprehensive DCaaS offering should include third-party vendor coordination, so that enterprises no longer need to deal with suppliers or schedule maintenance windows when parts need to be replaced or services need to be adjusted.
Digital transformation is a major strategic initiative for companies to undertake, and they should only entrust their core business infrastructure and operations to providers who can provide the assurance of guaranteed, measurable results. For example, Mirantis Flow, an open source cloud native Data Center as a Service, offers a number of guaranteed outcomes, such as up to 99.99% availability SLA and 48-hour turnaround time from when their deployment engineers gain access to customer hardware and when customers can spin up their first VM or create their first container.
Focusing on your core business
Enterprises beginning their digital transformation journey should carefully weigh the benefits and risks of keeping IT operations in-house, and consider DCaaS as a viable approach to increasing operational efficiency and boosting competitiveness for their business. From the dot com era to today, the Data Center as a Service definition may have changed, but its basic premise has stayed the same: data centers are difficult to set up and maintain, and businesses need help.
Whereas originally Data Center as a Service enabled enterprises to offload facility space, physical security, power, cooling, and installation tasks, nowadays an open source cloud native Data Center as a Service offers a more complete solution.
The Cloud Native Data Center as a Service provides a comprehensive, provider-agnostic, multi-platform solution — encompassing software, services and support — that enables enterprises to adopt a “zero-ops” approach to IT, unified across any infrastructure they choose.
